Tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL) is a typical member of the TNF family, and is a membrane protein participating in apoptosis. TRAIL is a protein consisting of 281 amino acids (SEQ ID NO:1: MAMMEVQGGPSLGQTCVLIVIFTVLLQSLCVAVTYVYFTNELKQMQDKYSKSGIACFL KEDDSYWDPNDEESMNSPCWQVKWQLRQLVRKMILRTSEETISTVQEKQQNISPLVRE RGPQRVAAHITGTRGRSNTLSSPNSKNEKALGRKINSWESSRSGHSFLSNLHLRNGELVI HEKGFYYIYSQTYFRFQEEIKENTKNDKQMVQYIYKYTSYPDPILLMKSARNSCWSKDA EYGLYSIYQGGIFELKENDRIFVSVTNEHLIDMDHEASFFGAFLVG), in which an extracellular domain comprising amino acids from arginine at position 114 to glycine at position 281 (SEQ ID NO:2: RERGPQRVAAHITGTRGRSNTLSSPNSKNEKALGRKINSWESSRSGHSFLSNLHLRNGEL VIHEKGFYYIYSQTYFRFQEEIKENTKNDKQMVQYIYKYTSYPDPILLMKSARNSCWSK DAEYGLYSIYQGGIFELKENDRIFVSVTNEHLIDMDHEASFFGAFLVG) affects apoptosis. Three molecules of TRAIL form a structurally modified trimer. The TRAIL trimer assembles with receptors participating in cell death to induce apoptosis.
A major difference between TRAIL and other members of the TNF superfamily, such as TNF and CD95L, is its ability not to induce cell death at normal tissues. A variety of medical or pharmaceutical applications have been attempted on TNF and CD95L, because these proteins induce cell death. Since TNF and CD95L proteins affect normal cells and also induce the death of cancer cells and over-activated immune cells, they have limited applicability. In contrast, TRAIL induces apoptosis in a wide range of cancer cells and over-activated immune cells with little effect on normal cells. This is due to the differential expression of TRAIL receptors between cell types. Five TRAIL receptors have been identified. Among them, DR4 (TRAIL-R1) and DR5 (TRAIL-R2) are representative cell death-related receptors. When TRAIL binds to DR4 or DR5, an intracellular death domain of the receptor is activated and thus transduces apoptotic signals via various signal transduction pathways, leading to apoptotic cell death. TRAIL can also bind to DcR1, DcR2 and osteoprotegerin (OPG), which do not induce apoptosis. No marked difference has been seen in the expression levels of the cell death-inducing receptors DR4 and DR5 between normal and tumor cells. In contrast, the three other receptors not inducing apoptosis are expressed at high levels in normal cells, but are either expressed at low levels or are not expressed at all in tumor cells. Thus, in normal cells, TRAIL binds mostly to DcR1, DcR2 and OPG, which do not contain a death domain, and thereby do not induce cell death. In contrast, in cancer cells and over-activated immune cells, apoptosis is induced by the binding of TRAIL to DR4 and DR5, which contain a death domain. Such selective apoptosis induction of TRAIL seems to be a particularly attractive feature in medical or pharmaceutical applications.
TRAIL-mediated apoptosis has been observed in various types of cancer cells, including colon carcinoma, glioma, lung carcinoma, prostate carcinoma, brain tumors and multiple myeloma cells. TRAIL has been proven to have very high anticancer activity in animals. The good anticancer efficacy of TRAIL has been obtained through the use of TRAIL alone, as well as in combination with other anticancer agents, such as paclitaxel and doxorubicin, and radiotherapy. Clinical trials are currently being conducted by Genentech and Amgen using TRAIL, which has good anticancer efficacy in solid tumors. In addition to cancer, various approaches using TRAIL have been made in arthritis, an autoimmune disease, for relieving and treating arthropathy by inducing the death of overactivated immune cells. In addition to protein therapy, gene therapy has been attempted through the delivery of the TRAIL gene. Thus, TRAIL may be useful in the treatment of various aforementioned types of cancer, as well as in the treatment of T cell-mediated autoimmune disorders such as experimental autoimmune encephalomyelitis, rheumatoid arthritis and type I diabetes.
However, native TRAIL has some problems to be overcome for application thereof. The major problem is the low trimer formation ratio of native TRAIL. TRAIL monomers do not bind to the aforementioned TRAIL receptors, and thus do not induce apoptosis. In this regard, many studies have been performed with the goal of improving the trimeric structure and trimer formation ratio of TRAIL. In native TRAIL, the zinc ion has been known to play a critical role in trimerization. In addition, the structural analysis of TRAIL has been conducted using a computer, and mutants of TRAIL have been developed based on the analysis results. For the formation of TRAIL trimers, the most useful method appears to be the introduction of a novel amino acid sequence favoring trimeric folding. Such sequences include a leucine zipper motif and an isoleucine zipper motif. Henning Walczak reported the anticancer efficacy of a trimeric TRAIL derivative in which a leucine zipper motif is added to the N terminus of native TRAIL (H. Walczak, et al. Nature Medicine 1999, 5, 157-163). Dai-Wu Seol et al. reported a TRAIL derivative containing a novel isoleucine zipper motif and having good apoptotic activity (M H Kim, et al. BBRC 2004, 321, 930-935).
Another problem in the clinical applications of TRAIL involves cytotoxicity shown in normal cells of some tissues. Most normal cells are resistant to cytotoxicity, resulting from the expression of the various aforementioned TRAIL receptors, but some hepatocytes and keratinocytes are sensitive to TRAIL-mediated cytotoxicity (H. Yagita, et al. Cancer Sci. 2004, 95, 777-783; M. Jo et al. Nature Medicine 2000, 6, 564-567; S. J. Zheng, et al. J. Clin. Invest. 2004, 113, 58-64).
In addition to toxicity toward some normal cells, TRAIL has a short half-life in vivo, which should be overcome for the successful clinical application of TRAIL. TRAIL has different half-lives according to the species of animals used in tests. For example, TRAIL has been reported to have a half-life of several minutes in rodents and about 30 minutes in apes (H. Xiang, et al. Drug Metabolism and Disposition 2004, 32, 1230-1238). In particular, most TRAIL is rapidly excreted via the kidneys. This short half-life is considered a drawback to the pharmaceutical usefulness of TRAIL, resulting in a need for TRAIL or derivatives thereof having an extended half-life. Other problems to be solved include the low solubility and solution stability of TRAIL.
Meanwhile, polyethylene glycol (PEG) is a polymer having a structure of HO—(—CH2CH2O—)n—H. Due to its high hydrophilicity, PEG enables an increase in the solubility of drug proteins when linked thereto. In addition, when suitably linked to a protein, PEG increases the molecular weight of the modified protein while maintaining major biological functions, such as enzyme activity and receptor binding, thereby reducing urinary excretion, protecting the protein from cells and antibodies recognizing exogenous antigens, and decreasing protein degradation by proteases. The molecular weight of PEG, capable of being linked to proteins, ranges from about 1,000 to 100,000. PEG having a molecular weight higher than 1,000 is known to have very low toxicity. PEG having a molecular weight between 1,000 and 6,000 is distributed widely throughout the entire body and is metabolized via the kidney. In particular, PEG having a molecular weight of 40,000 is distributed in the blood and organs, including the liver, and is metabolized in the liver.
In general, medically and pharmaceutically useful proteins administered via parenteral routes are disadvantageous in terms of being immunogenic in the body, being poorly water-soluble and being cleared from circulation within a short period of time. Many studies have been performed to overcome such problems. U.S. Pat. No. 4,179,337 mentions that, when used as therapeutics, pegylated proteins and enzymes have effects including reduced immunogenicity, increased solubility and extended in vivo residence time, which all are advantages of PEG. After this patent, a variety of efforts have been made to overcome the drawbacks of bioactive proteins through pegylation. An example is Veronese et al. pegylated ribonuclease and superoxide dismutase (Veronese et al., 1985). U.S. Pat. Nos. 4,766,106 and 4,917,888 describes the conjugation of proteins to a polymer including PEG to increase the water solubility thereof. In addition, U.S. Pat. No. 4,902,502 describes the conjugation of recombinant proteins to PEG or other polymers to reduce immunogenicity and extend circulating in vivo half-life.
However, despite the aforementioned advantages, there are some problems with protein pegylation, as follows. PEG is typically conjugated to a target protein through covalent bonding to one or more free lysine (Lys) residues. At this time, if PEG is bound to a region directly associated with protein activity among surface regions of the protein, the PEG-attached region loses biological functions, leading to decreased protein activity. Also, since the attachment of PEG to lysine residues mostly occurs in a random manner, various kinds of PEG-protein conjugates, corresponding to particular attachment sites, exist as a mixture. From this mixture, a desired conjugate is difficult to purify and isolate.
TRAIL also harbors problems similar to those caused by pegylation described above. Native TRAIL has eleven lysine residues, some of which participate in the interaction between TRAIL and its cognate receptors or are within an active site. Thus, the addition of polyethylene glycol molecules through the reaction with lysine residues may act as a very important inhibitory factor against the bioactivity of TRAIL. In the case of the TNF superfamily, several studies have noted that the bonding between lysine residues and polyethylene glycol molecules inhibits activation (Y. Yamamoto, et al. Nature Biotechnology 2003, 21, 546-552; H. Shibata et al. Clin. Cancer Res. 2004, 10, 8293-8300).
In this regard, the present inventors selectively attached PEG or a PEG derivative to an N-terminus of TRAIL. Such pegylation was found to reduce drug uptake and removal by hepatocytes and the hepatic reticuloendothelial system, leading to a decrease in TRAIL-mediated hepatoxicity, and remarkably increase the solubility and stability of TRAIL. Also, pegylation was found to improve pharmacokinetic profiles of a linked drug with long-term storage in various formulations, thereby reducing drug administration frequencies and allowing sustained duration of effects of the drug. In this way, the present inventors obtained an N-terminal modified PEG-TRAIL conjugate and found a method of preparing the conjugate, thereby leading to the present invention.